Lokakarya Komputasi di BATAN

Rabu, 6 Agustus 2008 kemarin, temen-temen mahasiswa jurusan Komputasi Statistik sempet maen sebentar ke BATAN (Badan Tenaga Nuklir Nasional) di kawasan Serpong. Waktu pamitan ke ibuk sempet diwanti-wanti “Hah, nuklir..?? Ati-ati yo, Nduk..!”. Hehehe…ibuk emang sayaaang banget, ya. Tapi setelah kujelaskan bahwa ni acara lokakarya komputasi dalam sains dan teknologi nuklir, bukan mau ikut bereksperimen di laboratorium nuklir mereka, ibuk pun akhirnya merelakan aku pergi. Ciee, kayak mo kemana aja gitu. Kujanjikan cerita nanti malam sebagai oleh-oleh, asal ga minta oleh-oleh rudal aja, hihi…

Ditemani Mr. Syukri, kami berangkat pake bus kampus. Kalau dipikir-pikir, setahun terakhir ini bus kampus makin sering ngorbit kemana-mana. Pak Odoy capek juga mungkin nyupirin gerombolan mahasiswa yang super cerewet kalo dah naek bus. Yang sabar ya, Pak..

Jam 6.30 pagi semua dah siap berangkat supaya ga terjebak macet. Bener aja, siangan dikit, jalan tol dah begitu puuadet. Apalagi di antara kami ga ada yang tahu persis lokasi gedungnya. Bakal seru nih karena pasti ada acara nyasar-nyasaran. Yang penting kan selalu sempet berphoto-photo ria. Duh, narcis..tetep jalan.

Sampai di tempat, masih sepi. Tapi berhubung kami ada belasan orang, tetep aja harus ngantri untuk ngisi daftar hadir. Kebetulan banget, mereka yang hadir di lokakarya ini dapat seminar kit yang salah satunya berupa tas tangan warna biru! Birunya itu matching, bahkan puuuersis banget dengan warna biru seragam yang kami pake. Jadi deh, kumat lagi, foto-foto dong! Sekalian name tagnya digantung juga di leher. Tertulis di situ kami sebagai “Peninjau”. weee…keren ya..(kumat! kumat!)

Masuk ke ruangan, ternyata Pak Mirza belum datang. Yup, salah satu dosen Komputasi kami diundang sebagai salah satu pembicara undangan di bagian awal pembukaan lokakarya ini. Penasaran dengan makalah yang beliau bawa? Nih, ada sedikit abstraksi yang saya cuplik dari sinopsis lokakarya ini. Begitu hampir dimulai, Pak Mirza masuk ruangan dan otomatis mau bergabung sama kami di barisan peserta. Langsung kuingetin, “Lho, Pak! Kok ga di depan? Di sini kan buat peserta.” Temen-temen yang lain juga bilang gitu. Tapi kayaknya ikatan suku emang ga bisa dipisahkan, hehe. Baru setelah seorang panitia yang menyilahkan Pak Mirza duduk di depan, beliau akhirnya mau juga. Acara pun dimulai, dan kepala-kepala kami berangsur-angsur mendidih. Sumpah! Keren banget ni acara!

Grid and Parallel Computing
Dr. Said Mirza Pahlevi, M.Eng.

Grid computing is an emerging technology that enables the sharing of wide variety resources (e.g., computing power and data storage capacity) distributed across different organizations. There are three characteristics that a grid must have, namely (1) coordinated resource sharing that are not subject to centralized control, (2) using standard, open, general-purpose protocols and interfaces and (3) delivering nontrivial qualities of service in term of throughput, response time, etc. [1]. The first poin comes from the inherent characteristics of a grid, that is, users and resources live within different control domains. The second point is important to establish resource sharing arrangements dynamically with an interested party and enabling general purpose services and tools. The third point is related to providing aggregated computing capabilities so that the utility of the combined system is significantly greater than that of the sum of its parts. Grid computing technology has been applied to computationally-intensive scientific problems, such as high energy and nuclear data analysis [2] and drug discovery [3]. Computationally-intensive problems are usually solved by using parallel computing. To create a parallel program, a programmer needs to identify work that can be done in parallel, partition the work and perhaps data among processes and manage data access, communication and synchronization. Message Passing Interface (MPI) [4] and GridRPC [5] are parallel programming models that are used to create parallel programs in a grid. The former uses a data parallel technique which is based on the Single Program Multiple Data (SPMD) model while the latter uses a task parallel technique which is based on the client/server model. The talk will explore the basic concept of grid computing and present some grid projects that deal with computationally-intensive problems. It also explains the basic concept of the two grid parallel programming paradigms.

Advertisements

~ by Ayuning Mazasupa on August 18, 2008.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

 
%d bloggers like this: